RBF-FNN Control And Its Application In Flexiblejoint Robots

Fuzzy Neural Network Intelligent Control technology is an important technology for robot control. With modern industry increasingly large, integrated and complex, it is particularly important for the research of fuzzy neural network in the field of intelligent control.In this paper, mainly study is dynamic surface control(DSC) algorithm based on RBF-FNN. Enhance the robustness of the algorithm by introducing the sliding mode control(SMC) in DSC based on RBF-FNN, and the algorithm is used to control flexible-joint robots. The main work is as follows:(1) An adaptive RBF fuzzy neural network dynamic surface control(RBF-FNN-DSC) algorithm with independent learning function for a class of nonaffine pure-feedback systems whose have unknown time-delay functions and perturbed uncertainties is proposed. In order to enhance the control effect for this type of uncertain lower-triangular nonlinear systems, this paper takes advantage of RBF-FNN’s excellent capabilities of self- learning and knowledge expressing, which combined with DSC method at the same time, and stability is proved by Lyapunov method. The simulation example is provided to illustrate the dynamic performance of the proposed scheme is better than the adaptive dynamic surface controller based on neural network(RBF-NN-DSC).(2) An adaptive RBF fuzzy neural network dynamic surface control algorithm with independent learning function based on sliding mode control(RBF-FNN-DSSMC) is proposed. Because RBF-FNN-DSC algorithm does not take full account of the first-order filter ’s accuracy and disturbance’s uncertainties, these factors have a significant impact on the robustness of the controller. So the DSC method is improved by introducing the SMC, and combined with RBF-FNN to weaken the chattering problem. Compared to the RBF-FNN-DSC algorithm, simulation results show that the controller has better robustness.(3) A specific application of RBF-FNN-DSSMC algorithm is given in flexible-joint robots control. The RBF-FNN-DSSMC algorithm is used as the flexible-joint robot control system with strong coupling, strong nonlinear and time-varying structure, and this closed- loop system to be semiglobally uniformly ultimately bounded which is proved by Lyapunov method, that is, having stability. Simulation results show that the performance of this type of robot control has better stability and robustness.